Rotomolding works better in a vacuum.

Evacuating a rotational mold before you make a part may sound like a strange idea, but it's also a simple way to achieve faster cure times and improved part properties. For this reason, rotomolders should give serious consideration to switching from the standard "vented" process in favor of the vacuum process.

Conceived years ago but apparently never developed commercially, the idea of air evacuation from a rotational mold by vacuum has shown significant benefits in tests by Equistar Chemicals. Vacuum hastens part densification during processing, leaving fewer air voids in the part and thereby enhancing mechanical properties. Limited tests to date yielded 17% shorter cycle times and a 58% boost in low-temperature impact strength.

Moreover, the vacuum-molded parts retain their improved properties even with continued heating after they reach maximum cure. This may make it possible to co-mold materials having dissimilar melting points without material degradation. It's likely that less oxygen exposure may reduce resin oxidation, which may also contribute to the benefits.

Get the bubbles out faster

Rotomolding cycle times are affected by how long it takes to eliminate air bubbles formed during the heating cycle. The part wall generally takes about half the heating cycle to form completely. The rest of the heating time is required to diffuse air bubbles trapped between the resin particles as they fuse. The presence of bubbles indicates an "undercured" condition. A part is considered adequately cured when no more bubbles are present. The air does not leave the part but rather diffuses into the melt.

We explored vacuum molding in response to a customer's request for help in raising production rates. Our premise was that starting with less air in the mold would reduce air bubbles produced during heating. Also, we expected that a vacuum in the mold would promote efficient diffusion of bubbles into the part wall. Both hypotheses proved correct.

Putting vacuum to the test

Our tests used Equistar's MP643-661, a 3.5 MI, 0.940-g/cc LLDPE, to mold 5-gal tanks at 650 F with a 4:1 spin ratio, which are standard processing conditions.

The technical center produced the 1/8-in.-thick tanks using cast-aluminum molds. The machine holds two molds at a time, which allowed us to make parts with the standard vented mold and a vacuum mold side by side under identical rotation, heating, and cooling conditions.

The vacuum test mold was charged, closed, and sealed with high-temperature caulk. Then a vacuum of 15 in. Hg (0.5 arm.) was drawn on the mold using a small vacuum pump. After achieving this level of vacuum, the valve on the mold was closed and the vacuum line was disconnected before the mold was put into the oven.

Negative pressure in the mold was retained through the end of the molding cycle. Typically just 3-5 in. Hg of vacuum was lost, which could be attributed to volatiles given off during the heating cycle.

Heating times of 7 to 14 min were used to make parts that ranged from undercured to overcured. The degree of cure was determined by physical inspection of the part for bubble presence and a measurement of low-temperature brittle impact (LTBI). The heating cycles were followed by a cooling cycle consisting of a 6-min air blow and a 4-min water spray. Since all parts were cooled the same, crystal growth or quenching were assumed to be identical.

Payoff in part performance

The vacuum-molded part had a peak LTBI value at 12 min heating time, when no bubbles were present upon visual inspection. Not surprisingly, the 12-min control part molded with standard venting had many bubbles in the part wall. Even at 14 min heating time, the control part had not yet reached maximum impact value. based on the sparse bubble count at 14 min, we projected that the control part would be fully cured at about 14.5 min. The estimated 2.5-min shorter cure time for the vacuum-molded part represents a 17% improvement. Figure 1 shows density measurements of molded parts, which correlate well with visual bubble counts.

LTBI data [ILLUSTRATION FOR FIGURE 2 OMITTED] clearly indicate that the vacuum-molded parts outperformed control parts from standard vented molds. An impact value of 51 ft-lb was obtained for the vacuum-molded part after only 12 min cure time, whereas the control part achieved a maximum of 43 ft-lb after 14 min.

Comparing LBTI of both parts after 12-min cure time makes the improvement more obvious. The fully cured vacuum-molded part is 57% more impact resistant than the undercured control part.

Vacuum molding may give better LBTI at full cure due to the reduced oxygen environment inside the evacuated mold during the heat cycle. Rheological tests showed no difference in the amount of cross linking between the two parts.

Tests showed no differences between the control and vacuum-molded parts in terms of tensile yield strength, heat deflection, flexural modulus, ESCR, and shrinkage.

Broader cure window

Unlike standard vented parts, the vacuum-molded samples showed an extended plateau after achieving full cure, with no drop-off in LTBI. Lower oxygen exposure is again the most likely reason. The extended "cure window" is important in that it allows materials of dissimilar melting points to be molded together without risking oxidative degradation of the lower-melting material.

So far, few or no negative tradeoffs have been discovered for the vacuum technology. Tests are ongoing, but it should prove applicable to a wide range of materials commonly used by rotomolders.